According to the PICOS, we developed inclusion criteria: 1) participants (P)—all the patients were diagnosed with localized PCa without infiltration or invasion of the prostate outside the envelope or adjacent organs, without lymph node metastasis, and with a follow-up period ≥3 months after radiotherapy. 2) Interventions (I) and comparisons (C): comparing the efficacy on sexual function of BT versus EBRT. 3) Outcomes (O): the indicators were the scores of the PCSI scale and the EPIC scale regarding sexual function. 4) Study design (S): prospective and retrospective studies (including cohort studies and case–control studies).

We excluded the following articles: 1) the literature without relevant indicators; 2) duplicate publications; 3) reviews; 4) animal experiments; and 5) conference abstracts.

We identified relevant studies by searching PubMed, Cochrane Library, Embase, China National Knowledge Infrastructure (CNKI), and Wanfang database up to March 4, 2021. The search was performed using MeSH terms, such as “Prostatic Neoplasms,” “Brachytherapy,” “Radiotherapy,” and “Erectile Dysfunction.” In addition, we scanned through the reference lists of included studies to find additional pertinent articles. We also contacted the corresponding author to acquire information if the research results were incomplete or could not be found.

Two authors independently screened the titles and abstracts of retrieved articles and then reviewed the full texts according to the inclusion and exclusion criteria. Then, the two authors extracted the data available in the included studies to fill out the well-designed form and check with each other. Any disagreement between the two authors was reviewed and resolved through a third author. For articles that only provided data such as the mean, sample size, and CI, RevMan calculator was used to convert them and calculate the required variables for the statistics. For articles that only provided X–Y scatter plots, WebPlotDigitizer was used to extract the data.

The Newcastle–Ottawa Scale (NOS) was used to evaluate the quality of the included studies from the following aspects, with the maximum score of 9 points: 1) representativeness of the exposed cohort; 2) selection of the nonexposed cohort; 3) ascertainment of exposure; 4) demonstration that the outcome of interest was not present at the time of initiating the study; 5) comparability of cohorts on the basis of the design or analysis; 6) assessment of the outcome; 7) the follow-up period long enough for the outcome to occur; and 8) adequacy of follow-up of cohorts. The studies with scores ≥6 were considered as high quality, and those with scores <6 were considered as low quality.

The risk of bias was assessed in four domains—selection bias, loss to follow-up bias, information bias, and confounding bias—based on which the risk of bias was classified as “low risk,” “unknown risk,” and “high risk.”

The meta-analysis was performed using the RevMan 5.4.1 software provided by the Cochrane Collaboration. Continuous data are presented as standardized mean difference (SMD) as effect size, with 95% CI calculated. Heterogeneity was evaluated by using the chi-square test and I² test. If there was no significant statistical difference (p > 0.05, I² < 50%), the fixed-effects model was used for analysis, and otherwise, the random-effects model was used. Heterogeneity was dealt with through subgroup analysis or sensitivity analysis. The means and SDs of the baseline were assumed to be X1 and S1, and the means and SD of the endpoint were X2 and S2. Then, we input “mean = X2 − X1” and “SD = S1 2 + S2 2 − 2 ∗ R ∗ S1 ∗ S2 (R = 0.5)” into the RevMan to make forest plots, which made the effect of the intervention mainly through the change in the amount of effect before and after the intervention, further eliminating the effect of the baseline.

A total of 962 citations were obtained through electronic databases, and 158 duplicates were eliminated by using EndNoteX9 software. After the titles and abstracts of the remaining 804 articles were screened, 792 articles we excluded due to irrelevancy. The full texts of the remaining 12 articles were reviewed, and finally, 8 articles were included for formal analysis according to the inclusion criteria, as detailed in Figure 1.

All the eight included studies were observational cohort studies, containing 2,340 patients who met the predefined inclusion criteria. All included studies involved the comparison of the effects of BT versus EBRT on sexual function, with 1,202 patients receiving EBRT and 1,138 patients receiving BT. The NOS scores of all the 8 articles were ≥6, with a mean score of 7.0. The basic characteristics are shown in Table 1.

The participants included in the 8 articles met the diagnostic criteria for localized PCa and had clear recorded medical histories regarding the treatment. In addition, the follow-up plan was defined appropriately before the evaluation of the sexual function and implemented as scheduled. Three studies (Ferrer et al., 2008; Chen et al., 2017; Mullins et al., 2019) independently evaluated the results using a blind method; four clinical trials (Ferrer et al., 2008; Guedea et al., 2009; van Tol-Geerdink et al., 2013; Chen et al., 2017) specified the missing rate at the beginning or described the lost visits after the trial; and five studies (Ferrer et al., 2008; Pardo et al., 2010; Ferrer et al., 2013; Jia and Zhang, 2018; Mullins et al., 2019) performed stratified or covariate analysis to control for confounding bias, as shown in Figures 2A,B, where “green” represents low risk, “yellow” represents unknown risk, and “red” represents high risk.

With the innovation of modern technology, the application of radiotherapy as one of the radical therapies for localized PCa has become increasingly mature. A prospective randomized controlled trial reported that there was no significant difference in cancer-specific survival (CSS) and 10-year OS between radiotherapy and radical surgery for localized PCa (Wallis et al., 2016). EBRT has evolved from traditional rotating irradiation and four-field box irradiation to intensity-modulated radiotherapy (IMRT), stereotactic radiotherapy (SBRT), and three-dimensional conformal radiotherapy (3D-CRT), which increases the dose to the target area but reduces the dose to surrounding normal tissues, thereby substantially reducing the occurrence of complications. BT is the implantation of radioactive particles into the human tissue for the purpose of radiation therapy, including temporary seed implantation and permanent seed implantation. The dose distribution around the radioactive particles is inversely proportional to the square of the distance from the radioactive source, ensuring a high dose to the local lesion tissue and a low dose to the surrounding normal tissue, thus reducing the damage to the surrounding normal tissue, which, together with the implantation of seeds to reduce the scope of surgical anatomy, provides further assurance of reducing impairment to the sexual function (de la Puente and Azab, 2014). Current evidence-based medicine indicates that BT alone is superior to EBRT alone in terms of efficacy and safety for the treatment of localized PCa (Kee et al., 2018; Li et al., 2018), but there is no conclusive evidence to define the difference in the effect of BT versus EBRT on the sexual function. The aim of the present study was to provide a basis for decision-making in the choice of clinical treatment.

After the integrated analysis of the eight articles included in this study, we found that the sexual function was less affected in patients with localized PCa who received BT as compared with that in patients who received EBRT in terms of the integrated scores of EPIC and PCSI. Although the overall heterogeneity was small (I² = 25%), subgroup analysis was carried out to make the study more convincing, because two different scales were adopted in the different studies included. The results showed less intragroup heterogeneity (I² = 24%) in the five studies using the EPIC scale (Ferrer et al., 2008; Guedea et al., 2009; van Tol-Geerdink et al., 2013; Ferrer et al., 2013; Pardo et al., 2010), and there was still a statistically significant difference between BT and EBRT. In contrast, analysis of the three studies using the PCSI scale (Chen et al., 2017; Jia and Zhang, 2018; Mullins et al., 2019) showed no intragroup heterogeneity (I² = 0%), but there was no significant difference in sexual function scores, and with higher heterogeneity between the two subgroups (I² = 58.2%). The result may be due to the smaller number of both studies using the PCSI scale and cases on the one hand; and on the other hand, it may be that the studies using the PCSI scale had a shorter mean follow-up period. Considering that the difference in radiation dose may have a certain influence on the results, we analyzed four studies (Ferrer et al., 2008; Guedea et al., 2009; Pardo et al., 2010; Ferrer et al., 2013) with the same radiation dose. The results revealed less intragroup heterogeneity (I² = 9%), and the difference between BT and EBRT was statistically significant (p = 0.004). This also greatly increased the persuasiveness of the article.

In view of the difference in the length of follow-up periods, we further carried out a stratified analysis and found that there was no statistically significant difference in the sexual function scores between the two methods at 3, 12, and 24 months after treatment. The reason may be that fewer studies were included in the stratified analysis, and most included studies used the PCSI scale. For example, four studies (Ferrer et al., 2008; Chen et al., 2017; Jia and Zhang, 2018; Mullins et al., 2019) were included in the 3-month analysis, of which three studies used the PCSI scale; five studies (Ferrer et al., 2008; van Tol-Geerdink et al., 2013; Chen et al., 2017; Jia and Zhang, 2018; Mullins et al., 2019) were included in the 12-month analysis, including three studies using the PCSI scale; and five studies (Ferrer et al., 2008; Guedea et al., 2009; Chen et al., 2017; Jia and Zhang, 2018; Mullins et al., 2019) were included in the 24-month analysis, of which three studies used the PCSI scale. In addition, the follow-up duration may not be long enough, and therefore longer-term follow-up studies are required to verify our findings and conclusions.

Most of the included articles were prospective cohort studies in the previous 10 years, with a large total number of cases. Although these studies used two different scales to evaluate the sexual function convincing, data analysis showed that there was low heterogeneity and the publication bias was within the acceptable limits, and sensitivity analysis showed that the results were stable, indicating that the conclusions of the present study are reliable and highly convincing.

Patients with PCa not only have high expectations for the therapeutic effect but are very much concerned about the changes in their quality of life. This study confirms that BT, to some extent, has a better therapeutic effect on the sexual function in patients than EBRT and therefore may provide a basis for the choice of clinical treatment.

First, this is an observational study, and the results obtained may be affected by various confounding factors. The lack of further grouping of tumors may also increase the study bias. Second, as the included articles were prospective cohort studies without randomized control trials, and the strength of the argument needs to be improved. Third, the original data of some studies were extracted by WebPlotDigitizer or calculated by related formulas and may produce errors in the results. Finally, this study only compared the effects of BT alone versus EBRT alone and did not include combination therapy (BT+EBRT) for comparison. Further in-depth research is required to verify our findings and conclusion.